U.S. Energy Information Administration - EIA - Independent Statistics and Analysis
Annual Energy Outlook 2013
International Executive Summary
Oil production, particularly from tight oil plays, rises over the next decade, leading to a reduction in net import dependence
Crude oil production has increased since 2008, reversing a decline that began in 1986. From 5.0 million barrels per day in 2008, U.S. crude oil production increased to 6.5 million barrels per day in 2012. Improvements in advanced crude oil production technologies continues to lift domestic supply, with domestic production of crude oil increasing in the Reference case before declining gradually beginning in 2020 for the remainder of the projection period. The projected growth results largely from a significant increase in onshore crude oil production, particularly from shale and other tight formations, which has been spurred by technological advances and relatively high oil prices. Tight oil development is still at an early stage, and the outlook is highly uncertain. In some of the AEO2013 alternative cases, tight oil production and total U.S. crude oil production are significantly above their levels in the Reference case.
The net import share of U.S. petroleum and other liquids consumption (including crude oil, petroleum liquids, and liquids derived from nonpetroleum sources) grew steadily from the mid-1980s to 2005 but has fallen in every year since then (Figure 1). In the Reference case, U.S. net imports of petroleum and other liquids decline through 2019, while still providing approximately one-third of total U.S. supply. The net import share of U.S. petroleum and other liquids consumption continues to decline in the Reference case, falling to 34 percent in 2019 before increasing to 37 percent in 2040.
The U.S. could become a net exporter of liquid fuels under certain conditions. An article in the Issues in focus section considers four cases that examine the impacts of various assumptions about U.S. dependence on imported liquids. Two cases (Low Oil and Gas Resource and High Oil and Gas Resource) vary only the supply assumptions, and two cases (Low/No Net Imports and High Net Imports) vary both the supply and demand assumptions. The different assumptions in the four cases generate wide variation from the liquid fuels import dependence values in the AEO2013 Reference case. In the Low/No Net Imports case, the United States ends its reliance on net imports of liquid fuels in the mid-2030s, with net exports rising to 8 percent of total U.S. liquid fuel production in 2040. In contrast, in the High Net Imports case, net petroleum import dependence is above 44 percent in 2040, which is higher than the Reference case level of 37 percent but still well below the 2005 level of 60 percent.
While other combinations of assumptions or unforeseen technology breakthroughs might produce a comparable outcome, the assumptions in the Low/No Imports case illustrate the magnitude and type of changes that would be required for the United States to end its reliance on net imports of liquid fuels, which began after World War II and has continued to the present day. Some of the assumptions in the Low/No Net Imports case, such as increased fuel economy for light-duty vehicles (LDVs) after 2025 and wider access to offshore resources, could be influenced by possible future energy policies. However, other assumptions in this case, such as the greater availability of onshore technically recoverable oil and natural gas resources, depend on geological outcomes that cannot be influenced by policy measures. In addition, economic trends, consumer preferences and behaviors, and technological factors also may be unaffected, or only modestly affected, by policy measures.
In the High Oil and Gas Resource case, changes due to the supply assumptions alone cause net import dependence to decline to 7 percent in 2040, with U.S. crude oil production rising to 10.2 million barrels per day in 2040, or 4.1 million barrels per day above the Reference case level. Tight oil production accounts for more than 77 percent (or 3.2 million barrels per day) of the difference in production between the two cases. Production of natural gas plant liquids in the United States also exceeds the Reference case level.
One of the most uncertain aspects of this analysis is the potential effect of different scenarios on the global market for liquid fuels, which is highly integrated. Strategic choices made by leading oil-exporting countries could result in U.S. price and quantity changes that differ significantly from those presented here. Moreover, regardless of how much the United States reduces its reliance on imported liquids, consumer prices will not be insulated from global oil prices if current policies and regulations remain in effect and world markets for delivery continue to be competitive.
U.S. dry natural gas production increases 1.3 percent per year throughout the Reference case projection, outpacing domestic consumption by 2019 and spurring net exports of natural gas (Figure 2). Higher volumes of shale gas production are central to higher total production volumes and a transition to net exports. As domestic supply has increased in recent years, natural gas prices have declined, making the United States a less attractive market for imported natural gas and more attractive for export.
U.S. net exports of natural gas grow to 3.6 trillion cubic feet in 2040 in the Reference case. Most of the projected growth in U.S. exports consists of pipeline exports to Mexico, which increase steadily as growing volumes of imported natural gas from the United States fill the widening gap between Mexico's production and consumption. Declining natural gas imports from Canada also contribute to the growth in U.S. net exports. Net U.S. imports of natural gas from Canada decline sharply from 2016 to 2022, then stabilize somewhat before dropping off again in the final years of the projection, as continued growth in domestic production mitigates the need for imports.
Continued low levels of liquefied natural gas (LNG) imports in the projection period, combined with increased U.S. exports of domestically sourced LNG, position the United States as a net exporter of LNG by 2016. U.S. exports of domestically sourced LNG (excluding exports from the existing Kenai facility in Alaska) begin in 2016 and rise to a level of 1.6 trillion cubic feet per year in 2027. One-half of the U.S. exports of LNG originate from the Lower 48 states and the other half from Alaska. The prospects for exports are highly uncertain, however, depending on many factors that are difficult to gauge, such as the development of new production capacity in foreign countries, particularly from deepwater reservoirs, shale gas deposits, and the Arctic. In addition, future U.S. exports of LNG depend on a number of other factors, including the speed and extent of price convergence in global natural gas markets and the extent to which natural gas competes with liquids in domestic and international markets.
In the High Oil and Gas Resource case, with more optimistic resource assumptions, U.S. LNG exports grow to more than 4 trillion cubic feet in 2040. Most of the additional exports originate from the Lower 48 states.
Natural gas consumption grows in industrial and electric power sectors as domestic production also serves an expanding export market
Relatively low natural gas prices, maintained by growing shale gas production, spur increased use in the industrial and electric power sectors, particularly over the next decade. In the Reference case, natural gas use in the industrial sector increases by 16 percent, from 6.8 trillion cubic feet per year in 2011 to 7.8 trillion cubic feet per year in 2025. After 2025, the growth of natural gas consumption in the industrial sector slows, while total U.S. consumption continues to grow (Figure 7). This additional growth is mostly for use in the electric power sector. Although natural gas continues to capture a growing share of total electricity generation, natural gas consumption by power plants does not increase as sharply as generation because new plants are very efficient (needing less fuel per unit of power output). The natural gas share of generation rose from 16 percent of generation in 2000 to 24 percent in 2011 and increases to 27 percent in 2025 and 30 percent in 2040. Natural gas use in the residential and commercial sectors remains nearly constant, as increasing end-use demand is balanced by increasing end-use efficiency.
Natural gas consumption also grows in other markets in the Reference case, including heavy-duty freight transportation (trucking) and as a feedstock for GTL production of diesel and other fuels. Those uses account for 6 percent of total U.S. natural gas consumption in 2040, as compared with almost nothing in 2011.
Natural gas use in the electric power sector grows even more sharply in the High Oil and Natural Gas Resource case, as the natural gas share of electricity generation grows to 39 percent, reaching 14.8 trillion cubic feet in 2040, more than 55 percent greater than in the Reference case. Industrial sector natural gas consumption growth is also stronger in this case, with growth continuing after 2025 and reaching 13.0 trillion cubic feet in 2040 (compared to 10.5 trillion cubic feet in 2040 in the Reference case). Much of the industrial growth in the High Oil and Natural Gas Resource case is associated with natural gas use for GTL production and increased lease and plant use in natural gas production.
International from Market Trends
Range of oil price cases represents uncertainty in world oil prices
In AEO2013, the Brent crude oil price is tracked as the main benchmark for world oil prices. The West Texas Intermediate (WTI) crude oil price has recently been discounted relative to other world benchmark crude prices. The recent growth in U.S. mid-continental oil production has exceeded the capacity of the oil transportation infrastructure out of Cushing, Oklahoma, the market center for WTI prices. The U.S. Energy Information Administration (EIA) expects the WTI price to approach levels near the Brent price as new oil pipeline capacity is added and begins operation.
Future oil prices are uncertain. EIA develops three oil price cases—Reference, High, and Low—to examine how alternative price paths could affect future energy markets (Figure 49). The AEO2013 price cases were developed by changing assumptions about four key factors: (1) the economics of petroleum liquids supply from countries outside the Organization of the Petroleum Exporting Countries (non-OPEC), (2) OPEC investment and production decisions, (3) the economics of other nonpetroleum liquids supply, and (4) world demand for petroleum and other liquids.
Relative to the Reference case, the Low Oil Price case assumes lower levels of world economic growth and liquid fuels demand, as well as more abundant and less costly non-OPEC liquid fuels supply. In the Low Oil Price case, OPEC supplies 49 percent of the worldâ€™s liquid fuels in 2040, compared with 43 percent in the Reference case. The High Oil Price case assumes higher levels of world economic growth and liquid fuels demand, along with less abundant and more costly non-OPEC liquid fuels supply. In the High Oil Price case, OPEC supplies 40 percent of the worldâ€™s liquid fuels in 2040.
Trends in petroleum and other liquids markets are defined largely by the developing nations
In the AEO2013 Reference, High Oil Price, and Low Oil Price cases, total world consumption of petroleum and other liquids in 2040 ranges from 111 to 118 million barrels per day (Figure 50). The alternative oil price cases reflect shifts in both supply and demand. Although demand at the margin in the Organization for Economic Cooperation and Development (OECD) countries is influenced primarily by price, demand in non-OECD regions, where future growth in world demand is concentrated, is driven primarily by rates of economic growth that are particularly uncertain. The AEO2013 Low Oil Price case reflects a scenario where slightly weaker economic growth limits non-OECD oil demand growth.
OECD petroleum and other liquids use grows in the Reference case to 47 million barrels per day in 2040, while non-OECD use grows to 65 million barrels per day. In the Low Oil Price case, OECD petroleum and other liquids use in 2040 is higher than in the Reference case, at 52 million barrels per day, but demand in the slow-growing non-OECD economies rises to only 59 million barrels per day. In the High Oil Price case, OECD consumption grows to 45 million barrels per day in 2040, and fast-growing non-OECD use—driven by higher GDP growth—increases to 73 million barrels per day in 2040.
The supply response also varies across the price cases. In the Low Oil Price case, OPEC's ability to manage its market share is weakened. Low prices have a negative impact on non-OPEC petroleum supply in comparison with the Reference case. In the High Oil Price case, OPEC restricts production, non-OPEC petroleum resources become more economical, and high oil prices make other liquids more economically attractive.
Production of liquid fuels from biomass, coal, and natural gas increases
In 2011, world production of liquid fuels from biomass, coal, and natural gas totaled 2.1 million barrels per day, or about 2 percent of the energy supplied by all liquid fuels. In the AEO2013 Reference case, production from the three sources grows to 5.7 million barrels per day in 2040 (Figure 51), or about 4 percent of the energy supplied by all liquid fuels.
In the Low Oil Price case, production of liquid fuels from these sources grows to 6.7 million barrels per day in 2040, as technology development is faster than projected in the Reference case, making the liquids easier to produce at lower cost, and demand for ethanol for use in existing blend ratios is higher. In the High Oil Price case, production grows to 9.1 million barrels per day in 2040, as higher prices stimulate greater investment in advanced liquid fuels technologies.
Across the three oil price cases, the largest contributions to production of advanced liquid fuels come from U.S. and Brazilian biofuels. In the Reference case, biofuel production totals 4.0 million barrels per day in 2040, and production of gas-to-liquids (GTL) and coal-to-liquids (CTL) fuels accounts for 1.7 million barrels per day of additional production in 2040. Biofuels production in 2040 totals 5.5 million barrels per day in the Low Oil Price case and 5.9 million barrels per day in the High Oil Price case. The projections for CTL and GTL production are more sensitive to world oil prices, varying from 1.2 million barrels per day in the Low Oil Price case to 3.3 million barrels per day in the High Oil Price case in 2040. In the Reference case, the U.S. share of world GTL production in 2040 is 36 percent, as recent developments in domestic shale gas supply have contributed to optimism about the long-term outlook for U.S. GTL plants.
With production outpacing consumption,
U.S. exports of natural gas exceed imports
The United States consumed more natural gas than it produced in 2011, with net imports of almost 2 trillion cubic feet. As domestic supply has increased, however, natural gas prices have declined, making the United States a less attractive market and reducing U.S. imports. Conversely, lower prices have made purchases of U.S. natural gas more attractive, increasing exports. In the AEO2013 Reference case, the United States becomes a net exporter of natural gas by 2020 (Figure 89).
Production growth, led by increased development of shale gas resources, outpaces consumption growth in the Reference case—a pattern that continues through 2040. As a result, exports continue to grow at a rate of about 17.7 percent per year from 2020 to 2040. Net exports in 2020 are less than 1 percent of total consumption; in 2040 they are 12 percent of consumption.
U.S. natural gas production increases by about 1 percent per year from 2011 to 2040 in the Reference case, meeting domestic demand while also allowing for more exports. The prospects for future exports are highly uncertain, however, depending on many factors that are difficult to anticipate, such as the development of new production capacity in foreign countries, particularly from deepwater reservoirs, shale gas deposits, and the Arctic.
Pipeline exports increase as Canadian imports fall and exports to Mexico rise
With relatively low natural gas prices in the AEO2013 Reference case, the United States becomes a net exporter of natural gas in 2020, and net exports grow to 3.6 trillion cubic feet in 2040 (Figure 92). Most of the projected growth in U.S. exports consists of pipeline exports to Mexico, which increase steadily over the projection period, as increasing volumes of imported natural gas from the United States fill the growing gap between Mexico's production and consumption. Exports to Mexico increase from 0.5 trillion cubic feet in 2011 to 2.4 trillion cubic feet in 2040.
U.S. exports of domestically sourced LNG (excluding existing exports from the Kenai facility in Alaska, which fall to zero in 2013) begin in 2016 and rise to a level of 1.6 trillion cubic feet per year in 2027. One-half of the projected increase in U.S. exports of LNG originate in the Lower 48 states and the other half from Alaska. Continued low levels of LNG imports through the projection period position the United States as a net exporter of LNG by 2016. In general, future U.S. exports of LNG depend on a number of factors that are difficult to anticipate, including the speed and extent of price convergence in global natural gas markets, the extent to which natural gas competes with oil in domestic and international markets, and the pace of natural gas supply growth outside the United States.
Net natural gas imports from Canada decline sharply from 2016 to 2022, then stabilize somewhat before dropping off again in the final years of the projection, as continued growth in domestic production mitigates the need for imports. Even as overall consumption exceeds supply in the United States, some natural gas imports from Canada continue, based on regional supply and demand conditions.
Domestic production of tight oil leads to lower imports of light sweet crude oil
API gravity is a measure of the specific gravity, or relative density, of a liquid, as defined by the American Petroleum Institute (API). It is expressed in degrees, where a higher number indicates lower density. Refineries generally process a mix of crude oils with a range of API gravities in order to optimize refinery operations. Over the past 15 years, the API gravity of crude oil processed in U.S. refineries has averaged between 30 and 31 degrees. As U.S. refiners run more domestic light crude produced from tight formations, they need less imported light oil crude to maintain an optimal API gravity. With increasing U.S. production of light crude oil in the Reference case, the average API gravity of crude oil imports declines (Figure 98).
In the AEO2013 Reference case, the trend toward increasing imports of heavier crude oils continues through 2035 before stabilizing . The increase in demand for diesel fuel in the projection, from 3.5 to 4.3 million barrels per day, leads to an increase in distillate and gas oil hydrocracking capacity (which increases diesel production capability) from 1.6 to 3.0 million barrels per day from 2011 to 2040.
The large increase in domestic production of light crude oil and the increase in imports of heavier crude oils have prompted significant investments in crude midstream infrastructure, including pipelines that will bring higher quantities of light sweet crudes to petroleum refineries along the U.S. Gulf Coast. In addition, significant investments are being made to move crude oil to refineries by rail. The Reference case assumes that sufficient infrastructure investments will be made through 2040 to move both light and heavy crude oils.
Increasing U.S. supply results in decreasing net imports of petroleum and other liquids
The net import share of U.S. petroleum and other liquids consumption (including crude oil, petroleum liquids, and liquids derived from nonpetroleum sources) grew steadily from the mid-1980s to 2005 but has fallen in every year since then. In the AEO2013 Reference and High Oil Price cases, U.S. imports of petroleum and other liquids decline through 2020, while still providing approximately one-third of total U.S. supply. As a result of increased production of domestic petroleum, primarily from tight oil formations, and a moderation of demand growth with tightening fuel efficiency standards, the import share of total supply declines. Domestic production of crude oil from tight oil formations, primarily from the Williston, Western Gulf, and Permian basins, increases by about 1.5 million barrels per day from 2011 to 2016 in both the Reference and High Oil Price cases.
The net import share of U.S. petroleum and other liquids consumption, which fell from 60 percent in 2005 to 45 percent in 2011, continues to decline in the Reference case, with the net import share falling to 34 percent in 2019 before increasing to 37 percent in 2040 (Figure 99). In the High Oil Price case, the net import share falls to an even lower 27 percent in 2040. In the Low Oil Price case, the net import share remains relatively flat in the near term but rises to 51 percent in 2040, as domestic demand increases, and imports become less expensive than domestically produced crude oil.
As a result of increased domestic production and slow growth in consumption, the United States becomes a net exporter of petroleum products, with net exports in the Reference case increasing from 0.3 million barrels per day in 2011 to 0.7 million barrels per day in 2040. In the High Oil Price case, net exports of petroleum products increase to 1.2 million barrels per day in 2040.
International from Issues in Focus
The benchmark oil price in AEO2013 is based on spot prices for Brent crude oil (commonly cited as Dated Brent in trade publications), an international benchmark for light sweet crude oil. The West Texas Intermediate (WTI) price has diverged from Brent and other benchmark prices over the past few years as a result of rapid growth in U.S. midcontinent and Canadian oil production, which has overwhelmed the transportation infrastructure needed to move crude oil from Cushing, Oklahoma, where WTI is quoted, to the Gulf Coast. EIA expects the WTI discount to the Brent price level to decrease over time as additional pipeline projects come on line, and will continue to report WTI prices (a critical reference point for the value of growing production in the U.S. midcontinent), as well as imported refiner acquisition costs (IRAC).
AEO2013 projections of future oil supply include two broad categories: petroleum liquids and other liquid fuels. The term petroleum liquids refers to crude oil and lease condensate—which includes tight oil, shale oil, extra-heavy crude oil, and bitumen (i.e., oil sands, either diluted or upgraded), plant condensate, natural gas plant liquids (NGPL), and refinery gain. The term other liquids refers to oil shale (i.e., kerogen-to-liquids), gas-to-liquids (GTL), coal-to-liquids (CTL), and biofuels (including biomass-to-liquids).
The key factors determining long-term supply, demand, and prices for petroleum and other liquids can be summarized in four broad categories: the economics of non-Organization of the Petroleum-Exporting Countries (OPEC) petroleum liquids supply; OPEC investment and production decisions; the economics of other liquids supply; and world demand for petroleum and other liquids.
To reflect the significant uncertainty associated with future oil prices, EIA develops three price cases that examine the potential impacts of different oil price paths on U.S. energy markets (Figure 21). The three price cases are developed by adjusting the four key factors described above. The following sections discuss the adjustments made in AEO2013. Each price case represents one of potentially many combinations of supply and demand that would result in the same price path. EIA does not assign probabilities to any of the oil price cases.
Because EIA's oil price paths represent market equilibrium between supply and demand in terms of annual average prices, they do not show the price volatility that occurs over days, months, or years. As a frame of reference, over the past two decades, volatility within a single year has averaged about 30 percent . Although that level of volatility could continue, the alternative oil price cases in AEO2013 assume smaller near-term price variation than in previous AEOs, because larger near-term price swings are expected to lead to market changes in supply or demand that would dampen the price.
The AEO2013 oil price cases represent internally consistent scenarios of world energy production, consumption, and economics. One interesting outcome of the three oil price cases is that, although the price paths diverge, interactions among the four key factors lead to nearly equal total volumes of world liquids supply in the three cases in the 2030 timeframe (Figure 22).
Among the key factors defining the Reference case are the Organization for Economic Cooperation and Development (OECD) and non-OECD gross domestic product (GDP) growth rates and liquid fuels consumption per dollar of GDP. Both the OECD and non-OECD growth rates and liquids fuels consumption per dollar of GDP decline over the projection period in the Reference case. OPEC continues restricting production in a manner that keeps its market share of total liquid fuels production between 39 percent and 43 percent for most of the projection, rising to 43 percent in the final years. Most other liquid fuels production technologies are economical at Reference case prices. In the Reference case, the Brent price declines to $96 per barrel in 2015 and then increases over the remainder of the period, to $163 per barrel in 2040, as a result of demand increases and supply pressures.
OPEC production in the Reference case grows from 35 million barrels per day in 2011 to 48 million barrels per day in 2040 (Figure 23). Although the OPEC resource base is sufficient to support much higher production levels, the OPEC countries have an incentive to restrict production in order to support higher prices and sustain revenues in the long term. The Reference case assumes that OPEC will maintain a cohesive policy of limiting supply growth, rather than maximizing total annual revenues. The Reference case also assumes that no geopolitical events will cause prolonged supply shocks in the OPEC countries that could further limit production growth.
Non-OPEC petroleum production grows significantly in the early years of the Reference case projection, to 55 million barrels per day in 2020 from 50 million barrels per day in 2011, primarily as a result of increased production from tight oil formations. After 2020, production growth continues at a slower pace, adding another 4 million barrels per day to net production in 2040, with production from new wells increasing slightly faster than the decline in production from existing wells. The growth in non-OPEC production results primarily from the development of new fields and the application of new technologies, such as enhanced oil recovery (EOR), horizontal drilling, and hydraulic fracturing, which increase recovery rates from existing fields. The average cost per barrel of non-OPEC oil production rises as production volumes increase, and the rising costs dampen further production growth.
Non-OPEC production of other liquids grows from 1.8 million barrels per day in 2011 to 4.6 million barrels per day in 2040, as Brent crude oil prices remain sufficiently high to make other liquids production technologies economically feasible. Non-OPEC liquids production in the Reference case totals 58 million barrels per day in 2020, 61 million barrels per day in 2030, and 64 million barrels per day in 2040.
Low Oil Price case
The AEO2013 Low Oil Price case assumes slower GDP growth for the non-OECD countries than in the Reference case. OPEC is less successful in restricting production in the Low Oil Price case, and as a result its share of total world liquids production increases to 49 percent in 2040. Despite lower Brent prices than in the Reference case, non-OPEC petroleum production levels are maintained at roughly 54 million barrels per day through 2030. After 2030, total non-OPEC production declines as existing fields are depleted and not fully replaced by production from new fields and more costly EOR technologies. With higher average costs for resource development in the non-OPEC countries, the Brent crude oil price in the Low Oil Price case is not sufficient to make all undeveloped fields economically viable. Non-OPEC petroleum production rises slightly in the projection, to 54 million barrels per day, before returning to roughly current levels of 51 million barrels per day in 2040. Non-OPEC production of other liquids grows more rapidly than in the Reference case, and in 2040 it is 25 percent higher than projected in the Reference case.
Brent crude oil prices fall below $80 per barrel in 2015 in the Low Oil Price case and decline further to just below $70 per barrel in 2017, followed by a slow increase to $75 per barrel in 2040. In the near term, extra supply enters the market, and lower economic growth in the non-OECD countries leads to falling prices. The higher levels of OPEC petroleum production assumed in the Low Oil Price case keep prices from increasing appreciably in the long term.
OPEC's ability to support higher oil prices is weakened by its inability to limit production as much as in the Reference case. Lower prices squeeze the revenues of OPEC members, increasing their incentive to produce beyond their quotas. As a result, OPEC liquids production increases to 54 million barrels per day in 2040. The lower prices in the Low Oil Price case cause a decline in OPEC revenue to the lowest level among the three cases, illustrating the relatively strong incentive for OPEC members to restrict supply.
High Oil Price case
In the High Oil Price case, non-OECD GDP growth is more rapid than projected in the Reference case, and liquid fuels consumption per unit of GDP in the non-OECD countries declines more slowly than in the Reference case. Continuing restrictions on oil production keep the OPEC market share of total liquid fuels production between 37 and 40 percent, with total oil production about 1.0 million barrels per day lower than in the Reference case. Despite higher Brent oil prices, non-OPEC petroleum production initially expands at about the same rate as in the Reference case because of limited access to existing resources and lower discovery rates. Non-OPEC production of other liquids grows strongly in response to higher prices, rising to 8 million barrels per day in 2040.
Brent crude oil prices in the High Oil Price case increase to $155 per barrel in 2020 and $237 per barrel in 2040 in reaction to very high demand for liquid fuels in the non-OECD countries. The robust price increase keeps total world demand within the range of expected production capabilities.
Liquid fuels  play a vital role in the U.S. energy system and economy, and access to affordable liquid fuels has contributed to the nation's economic prosperity. However, the extent of U.S. reliance on imported oil has often been raised as a matter of concern over the past 40 years. U.S. net imports of petroleum and other liquid fuels as a share of consumption have been one of the most watched indicators in national and global energy analyses. After rising steadily from 1950 to 1977, when it reached 47 percent by the most comprehensive measure, U.S. net import dependence declined to 27 percent in 1985. Between 1985 and 2005, net imports of liquid fuels as a share of consumption again rose, reaching 60 percent in 2005. Since that time, however, the trend toward growing U.S. dependence on liquid fuels imports has again reversed, with the net import share falling to an estimated 41 percent in 2012, and with EIA projecting further significant declines in 2013 and 2014. The decline in net import dependence since 2005 has resulted from several disparate factors, and continued changes in those and other factors will determine how this indicator evolves in the future. Key questions include:
- What are the key determinants of U.S. liquid fuels supply and demand?
- Will the supply and demand trends that have reduced dependence on net imports since 2005 intensify or abate?
- What supply and demand developments could yield an outcome in which the United States is no longer a net importer of liquid fuels?
This discussion considers potential changes to the U.S. energy system that are inherently speculative and should be viewed as what-if cases. The four cases that are discussed include two cases (Low Oil and Gas Resources and High Oil and Gas Resources) in which only the supply assumptions are varied, and two cases (Low/No Net Imports and High Net Imports) in which both supply and demand assumptions change. The changes in these cases generate wide variation from the liquid fuels import dependence values seen in the AEO2013 Reference case, but they should not be viewed as spanning the range of possible outcomes. Cases in which both supply and demand assumptions are modified show the greatest changes. In the Low/No Net Imports case, the United States ceases to be a net liquid fuels importer in the mid-2030s, and by 2040 U.S. net exports are 8 percent of total U.S. liquid fuel production. In contrast, in the High Net Imports case, net petroleum import dependence is above 44 percent in 2040, higher than the Reference case level of 37 percent but still well below the 60-percent level seen in 2005. Cases in which only supply assumptions are varied show intermediate levels of change in liquid fuels import dependence.
As the case names suggest, the Low Oil and Gas Resource case incorporates less-optimistic oil and natural gas resource assumptions than those in the Reference case, while the High Oil and Gas Resource case does the opposite. The other two cases combine different oil and natural gas resource assumptions with changes in assumptions that influence the demands for liquid fuels. The Low/No Net Imports case simulates an environment in which U.S. energy production grows rapidly while domestic consumption of liquid fuels declines. Conversely, the High Net Imports case combines the Low Oil and Gas Resource case assumptions with demand-related assumptions including slower improvements in vehicle efficiency, higher levels of vehicle miles traveled (VMT) relative to the Reference case, and reduced use of alternative transportation fuels.
A key contributing factor to the recent decline in net import dependence has been the rapid growth of U.S. oil production from tight onshore formations, which has followed closely after the rapid growth of natural gas production from similar types of resources. Projections of future production trends inevitably reflect many uncertainties regarding the actual level of resources available, the difficulty in extracting them, and the evolution of the technologies (and associated costs) used to recover them. To represent these uncertainties, the assumptions used in the High and Low Oil and Gas Resource cases represent significant deviations from the Reference case.
Estimates of technically recoverable resources from the rapidly developing tight oil formations are particularly uncertain and change over time as new information is gained through drilling, production, and technology experimentation. Over the past decade, as more tight and shale formations have gone into commercial production, estimates of technically and economically recoverable resources have generally increased. Technically recoverable resource estimates, however, embody many assumptions that might not prove to be true over the long term, over the entire range of tight or shale formations, or even within particular formations. For example, the tight oil resource estimates in the Reference case assume that production rates achieved in a limited portion of a given formation are representative of the entire formation, even though neighboring tight oil well production rates can vary widely. Any specific tight or shale formation can vary significantly across the formation with respect to relevant characteristics , resulting in widely varying rates of well production. The application of refinements to current technologies, as well as new technological advancements, can also have a significant but highly uncertain impact on the recoverability of tight and shale crude oil.
As shown in Table 5, the High and Low Oil and Gas Resource cases were developed with alternative crude oil and natural gas resource assumptions giving higher and lower technically recoverable resources than assumed in the Reference case. While these cases do not represent upper and lower bounds on future domestic oil and natural gas supply, they allow for an examination of the potential effects of higher and lower domestic supply on energy demand, imports, and prices.
The Low Oil and Gas Resource case only reflects the uncertainty around tight oil and shale gas resources. The resource estimates in the Reference case are based on crude oil and natural gas production rates achieved in a limited portion of the tight or shale formation and are assumed to be representative of the entire formation. However, the variability in formation characteristics described earlier can also affect the estimated ultimate recovery (EUR) of wells. For the Low Oil and Gas Resource case, the EUR per tight and shale well is assumed to be 50 percent lower than in the AEO2013 Reference case. All other resource assumptions are unchanged from the Reference case.
The High Oil and Gas Resource case reflects a broad-based increase in crude oil and natural gas resources. Optimism regarding increased supply has been buoyed by recent advances in crude oil and natural gas production that resulted in an unprecedented annual increase in U.S. crude oil production in 2012. The AEO2013 Reference case shows continued near-term production growth followed by a decline in U.S. production after 2020. The High Oil and Gas Resource case presents a scenario in which U.S. crude oil production continues to expand after about 2020 due to assumed higher technically recoverable tight oil resources, as well as undiscovered resources in Alaska and the offshore Lower 48 states. In addition, the maximum annual penetration rate for GTL technology is doubled compared to the Reference case.
The tight and shale resources are increased by changing both the EUR per well and the well spacing. A doubling in tight and shale well EUR, when assumed to occur through raising the production type curves  across the board, is responsible for the significantly faster increases in production and is also a contributing factor in avoiding the production decline during the projection period. This assumption change is quite optimistic and may alternatively be considered as a proxy for other changes or combinations of changes that have yet to be observed.
Although initial production rates have increased over the past few years, it is too early to conclude that overall EURs have increased and will continue to increase. Instead, producers may just be recovering the resource more quickly, resulting in a more dramatic decline in production later, with little impact on the well's overall EUR. The decreased well spacing reflects less the capability to drill wells closer together (i.e., avoid interference) and instead more the discovery of and production from other shale plays that are not yet in commercial development. These may either be stacked in the same formation or reflect future technological innovations that would bring into production plays that are otherwise not amenable to current hydraulic fracturing technology.
Other resources also are assumed to contribute to supply, as technological or other unforeseen changes improve their prospects. The resource assumptions for the offshore Lower 48 states in the High Oil and Gas Resource case reflect the possibility that resources may be substantially higher than assumed in the Reference case. Resource estimates for most of the U.S. Outer Continental Shelf are uncertain, particularly for resources in undeveloped regions where there has been little or no exploration and development activity, and where modern seismic survey data are lacking . The increase in crude oil resources in Alaska reflects the possibility that there may be more crude oil on the North Slope, including tight oil. It does not, however, reflect an opening of the Arctic National Wildlife Refuge to exploration or production activity. Finally, modest production from kerogen (oil shale) resources, which remains below 140,000 barrels per day through the 2040 projection horizon, is included in the High Oil and Gas Resource case.
Reductions in demand for liquid fuels in some uses, such as personal transportation and home heating, coupled with slow growth in other applications, have been another key contributing factor in the decline of the nation's net dependence on imported liquid fuels since 2005. As with supply assumptions, the key analytic assumptions that drive future trends in liquid fuels demand in EIA's projections are subject to considerable uncertainty. The most important assumptions affecting future demand for liquids fuels include:
- The future level of activities that use liquid fuels, such as VMT
- The future efficiency of equipment that uses liquid fuels, such as automobiles, trucks, and aircraft
- The future extent of fuel switching that replaces liquid fuels with other fuel types, such as liquefied natural gas (LNG), biofuels, or electricity.
Two alternative sets of demand assumptions that lead to higher or lower demand for liquid fuels than in the AEO2013 Reference case are outlined below. The two alternative scenarios are then applied in conjunction with the High and Low Oil and Gas Resource cases to develop the Low/No Net Import and High Net Import cases.
Vehicle miles traveled
Projected fuel use by LDVs is directly proportional to light-duty VMT, which can be influenced by policy, but it is driven primarily by market factors, demography, and consumer preferences. All else being equal, VMT is more likely to grow when the driving-age population is growing, economic activity is robust, and fuel prices are moderate. For example, there is a strong linkage between economic activity, employment, and commuting. In addition, there is a correlation between income and discretionary travel that reinforces the economy-VMT link. Turning to demography, factors such as the population level, age distribution, and household composition are perhaps most important for VMT. For example, lower immigration would lead to a smaller U.S. population over time, lowering VMT. The aging of the U.S. population continues and will also have long-term effects on VMT trends, as older drivers do not behave in the same ways as younger or middle-aged drivers. At times, the factors that influence VMT intertwine in ways that change long-term trends in U.S. driving and fuel consumption. For example, the increase in two-income families that occurred beginning in the 1970s created a surge in VMT that involved both economic activity and demographics.
Alternative modes of travel affect VMT to the degree that the population substitutes other travel services for personal LDVs. The level of change is related to the cost, convenience, and geographic extent of mass transit, rail, biking, and pedestrian travel service options. Car-sharing services, which have grown in popularity in recent years, could discourage personal vehicle VMT by putting more of the cost of incremental vehicle use on the margin when compared with traditional vehicle ownership or leasing, where many of the major costs of vehicle use are incurred at the time a vehicle is acquired, registered, and insured. Improvements in the fuel efficiency of vehicles, however, could increase VMT by lowering the marginal costs of driving. In recent analyses supporting the promulgation of new final fuel economy and GHG standards for LDVs in MY 2017 through 2025, NHTSA and EPA applied a 10-percent rebound in travel to reflect the lower fueling costs of more efficient vehicles . Both higher and lower values for the rebound have been advanced by various analysts .
Other types of technological change also can affect projected VMT growth. E-commerce, telework, and social media can supplant (or complement) personal vehicle use. Some analysts have suggested an association between rising interest in social media and a decline in the rates at which driving-age youth secure driver licenses; however, that decline also could be related to recent weakness in the economy.
Many of the factors reviewed above were also addressed in the August 2012 National Petroleum Council Future Transportation Fuels study . That study considered numerous specific research efforts, as well as available summaries of the literature on VMT, and concluded that the economic and demographic factors remain dominant. The VMT scenario adopted for most of the analysis in that study reflected declining compound annual growth rates of VMT over time, with the growth rate in VMT, which was 3.1 percent in the 1971-1995 and 2.0 percent in the 1996-2007 periods, falling to under 1 percent after 2035.
In the AEO2013 Reference case, the compound annual rate of growth in light-duty VMT over the period from 2011 to 2040 is 1.2 percent—well below the historical record through 2005 but significantly higher than the average annual light-duty VMT growth rate of 0.7 percent from 2005 through 2011. The 2005-2011 period was marked by generally poor economic performance, high unemployment, and high liquid fuel prices, all of which likely contributed to lower VMT growth. While VMT growth rates are expected to rise as the economy and employment levels improve, it remains to be seen to what extent such effects might be counteracted or reinforced by some of the other market factors identified above.
The low demand scenario used in the Low/No Net Imports case holds the growth rate of light-duty VMT over the 2011-2040 period at 0.2 percent per year, lower than its 2005-2011 growth rate. The application of a lower growth rate over a 29-year projection period results in total light-duty VMT 26 percent below the Reference case level in 2040. With population growth at 0.9 percent per year, this implies a decline of 0.7 percent per year in VMT per capita. VMT per licensed driver, which increases by 0.3 percent per year in the AEO2013 Reference case, declines at a rate of 0.8 percent per year in the Low/No Net Imports case. In the High Net Imports case, which assumes more robust demand than in the Reference case, the VMT projection remains close to that in the Reference case, with higher demand resulting from other factors.
Turning to vehicle efficiency, the rising fuel economy of new LDVs already has contributed to recent trends in liquid fuels use. Looking forward, the EPA and NHTSA have established joint CAFE and GHG emissions standards through MY 2025. The new CAFE standards result in a fuel economy, measured as a program compliance value, of 47.3 mpg for new LDVs in 2025, based on the distribution of production of passenger cars and light trucks by footprint in AEO2013. The EPA and NHTSA also have established a fuel efficiency and GHG emissions program for medium- and heavy-duty vehicles for MY 2014-18. The fuel consumption standards for MY 2014-15 set by NHTSA are voluntary, while the standards for MY 2016 and beyond are mandatory, except those for diesel engines, which are mandatory starting in 2017.
The AEO2013 Reference case does not consider any possible reduction in fuel economy standards resulting from the scheduled midterm review of the CAFE standards for MY 2023-25, or for any increase in fuel economy standards that may be put in place for model years beyond 2025. The low demand scenario in this article adopts the assumption that post-2025 LDV CAFE standards increase at an average annual rate of 1.4 percent, the same assumption made in the AEO2013 Extended Policies case. In contrast, the high demand scenario assumes some reduction in current CAFE standards following the scheduled midterm review.
In the AEO2013 Reference case, fuel switching to natural gas in the form of compressed natural gas (CNG) and LNG already is projected to achieve significant penetration of natural gas as a fuel for heavy-duty trucks. In the Reference case, natural gas use in heavy-duty vehicles increases to 1 trillion cubic feet per year in 2040, displacing 0.5 million barrels per day of diesel use. The use of natural gas in the Reference case is economically driven. Even after the substantial costs of liquefaction or compression, fuel costs for LNG or CNG are expected to be well below the projected cost of diesel fuel on an energy-equivalent basis. The fuel cost advantage is expected to be large enough in the view of a significant number of operators to offset the considerably higher acquisition costs of vehicles equipped to use these fuels, in addition to offsetting other disadvantages, such as reduced maximum range without refueling, a lower number of refueling locations, reduced volume capacity in certain applications, and an uncertain resale market for vehicles using alternative fuels. For purposes of the low demand scenario for liquid fuels, factors limiting the use of natural gas in heavy-duty vehicles are assumed to be less significant, allowing for higher rates of market penetration.
Natural gas could also prove to be an attractive fuel in other transportation applications. The use of LNG as a fuel for rail transport, which had earlier been considered for environmental reasons, is now under active consideration by major U.S. railroads for economic reasons, motivated by the same gap between the cost of diesel fuel and LNG now and over the projection period. Because all modern railroad locomotives use electric motors to drive their wheels, a switch from diesel to LNG would entail the use of a different fuel to drive the onboard electric generation system. Retrofits have been demonstrated, but new locomotives with generating units specifically optimized for LNG could prove to be more attractive. Because railroads already maintain their own on-system refueling infrastructure, they may be less subject to the concern that truckers considering a switch to alternative fuel vehicles might have regarding the risks that natural gas refueling systems they require would not actually be built. The high concentration of ownership in the U.S. railroad industry could also facilitate a rapid switch toward LNG refueling, with the associated transition to new equipment, under the right circumstances because there are only a few owners making the decisions.
Marine operators have traditionally relied on oil-based fuels, with large oceangoing vessels almost exclusively fueled with heavy high-sulfur fuel oil that typically sells at a discount relative to other petroleum products. Under the International Maritime Organization's International Convention on the Prevention of Pollution from Ships agreement (MARPOL Annex VI) , the use of heavy high-sulfur fuel oil in international shipping started being phased out for environmental reasons in 2010. Although LNG is one possible option, there are many cost and logistical challenges, including the high cost of retrofits, the long lifetime of existing vessels, and relatively low utilization rates for many routes that will have adverse impacts on the economics of marine LNG refueling infrastructure. Unlike the heavy-duty truck market, there has not yet been an LNG-fueled product offered for general use by manufacturers of marine or rail equipment, making cost and performance comparisons inherently speculative.
In addition to the demand assumptions discussed above, other assumption changes were made to capture potential shifts in vehicle cost and consumer preference for LDVs powered by alternative fuels. In the Low/No Net Imports case, the costs of efficiency technologies and battery technologies were lowered, and the market penetration of E85 fuel was increased, relative to the Reference case levels. With regard to E85, assumptions about consumer preference for flex-fuel vehicles were altered to allow for increases in vehicle sales and E85 demand, leading to greater use of domestically-produced biofuel than projected in the Reference case.
Table 6 summarizes the demand-side assumptions in the alternative demand scenarios for liquid fuels. As with the supply assumptions, the assumptions used in the higher and lower demand cases represent substantial deviations from the AEO2013 Reference case, and they might instead be realized in terms of other, as-yet-unforeseen developments in technology, economics, or policy.
The cases considered show how the future share of net imports in total U.S. liquid fuel use varies with changes in assumptions about the key factors that drive domestic supply and demand for liquid fuels (Figure 24). Some of the assumptions in the Low/No Net imports case, such as assumed increases in LDV fuel economy after 2025 and access to offshore resources, could be influenced by future energy policies. However, other assumptions in this case, such as the greater availability of onshore technically recoverable oil and natural gas resources, depend on geological outcomes that cannot be influenced by policy measures; and economic, consumer, or technological factors may likewise be unaffected or only slightly affected by policy measures.
Net imports and prices
In the Low/No Net Imports case, U.S. net imports of liquid fuels are eliminated in the mid-2030s, and the United States becomes a modest net exporter of those fuels by 2040. As discussed above, this case combines optimistic assumptions about the availability of domestic oil and natural gas resources with assumptions that lower demand for liquid fuels, including a decline in VMT per capita, increased switching to natural gas fuels for transportation (including heavy-duty trucks, rail, boats, and ships), continued significant improvements in the fuel efficiency of new vehicles beyond 2025, wider availability and lower costs of electric battery technologies, and greater market penetration of biofuels and other nonpetroleum liquids. Although other combinations of assumptions, or unforeseen technology breakthroughs, might produce a comparable outcome, the assumptions in the Low/No Net Imports case illustrate the magnitude and type of changes that would be required for the United States to end its reliance on net imports of liquid fuels, which began in 1946 and has continued to the present day. Moreover, regardless of how much the United States is able to reduce its reliance on imported liquids, it will not be entirely insulated from price shocks that affect the global oil market .
As shown in Figure 24, the supply assumptions of the High Oil and Gas Resource case alone result in a decline in net import dependence to 7 percent in 2040, compared to 37 percent in the Reference case, with U.S. crude oil production rising to 10.2 million barrels per day in 2040, or 4.1 million barrels per day above the Reference case level. Tight oil production accounts for more than 77 percent (or 3 million barrels per day) of the difference in production between the two cases. Production of NGL in the United States also exceeds the Reference case level.
As a result of higher U.S. liquid fuels production, Brent crude oil prices in the High Oil and Gas Resource case are lower than in the Reference case, which also lowers motor gasoline and diesel prices to the transportation sector, encouraging greater consumption and partially dampening the projected decline in net dependence on liquid fuel imports. In the High Oil and Gas Resource case, the reduction in motor fuels prices increases fuel consumption in 2040 by 350 thousand barrels per day in the transportation sector and 230 thousand barrels per day in the industrial sector, which accounts for nearly all of the increase in total U.S. liquid fuels consumption (600 thousand barrels per day) relative to the Reference case total in 2040.
Global market, the economy, and refining
The addition of assumptions that slow the growth of demand for liquid fuels in the Low/No Net Imports case more than offsets the increase in demand that results from lower liquid fuel prices, so that total liquid fuels consumption in 2040 is 2.1 million barrels per day lower than projected in the Reference case. The combination of high crude oil and natural gas resources and lower demand for liquid fuels pushes Brent crude oil prices to $29 per barrel below the Reference case level in 2040. However, given the cumulative impact of factors that tend to raise world oil prices in real terms over the projection period, inflation-adjusted crude oil prices in the Low/No Net Imports case are still above today's price level.
One of the most uncertain aspects of the analysis concerns the effect on the global market for liquid fuels, which is highly integrated. Although the analysis reflects price effects that are based on the relative scale of the changes in U.S. domestic supply and net U.S. imports of liquid fuels within the overall international crude oil market, strategic choices made by the leading oil-exporting countries could result in price and quantity effects that differ significantly from those presented here. Moreover, regardless of how much the United States reduces its reliance on imported liquids, consumer prices will not be insulated from global oil prices if current policies and regulations remain in effect and world markets for crude oil streams of sulfur quality remain closely aligned absent transportation bottlenecks .
Although the focus is mainly on liquid fuels markets, the more optimistic resource assumptions in the High Oil and Gas Resource case also lead to more natural gas production. The higher productivity of shale and tight gas wells puts downward pressure on natural gas prices and thus encourages increased domestic consumption of natural gas (38 trillion cubic feet in the High Oil and Gas Resource case, compared to 30 trillion cubic feet in the Reference case in 2040) and higher net exports (both pipeline and LNG) of natural gas. As a result, projected domestic natural gas production in 2040 is considerably higher in the High Oil and Gas Resource case (45 trillion cubic feet) than in the Reference case (33 trillion cubic feet).
The Low Oil and Gas Resource case illustrates the implications of an outcome in which U.S. oil and gas resources turn out to be smaller than expected in the Reference case. In this case, domestic crude oil production peaks in 2016 at 6.9 million barrels per day, declines to 5.9 million barrels per day in 2028, and remains relatively flat (between 5.8 and 6.0 million barrels per day) through 2040. The lower well productivity in this case puts upward pressure on natural gas prices, resulting in lower natural gas consumption and production. In 2040, U.S. natural gas production is 27 trillion cubic feet in the Low Oil and Gas Resource case, compared with 33 trillion cubic feet in the Reference case.
These alternative cases may also have significant implications for the broader economy. Liquid fuels provide power and raw materials (feedstocks) for a substantial portion of the U.S. economy, and the macroeconomic impacts of both the High Oil and Gas Resource case and the Low/No Net Imports case suggest that significant economic benefits would accrue if some version of those futures were realized (see discussion of NGL later in "Issues in focus"). This is in spite of the fact that petroleum remains a global market in each of the scenarios, which limits the price impacts for gasoline, diesel, and other petroleum-derived fuels. In the High Oil and Gas Resource case, increasing energy production has immediate benefits for the economy. U.S. industries produce more goods with 12 percent lower energy costs in 2025 and 15 percent lower energy costs in 2040. Consumers see roughly 10 percent lower energy prices in 2025, and 13 percent lower energy prices in 2040, as compared with the Reference case. Cheaper energy allows the economy to expand further, with real GDP attaining levels that are on average about 1 percent above those in the Reference case from 2025 through 2040, including growth in both aggregate consumption and investment.
The alternative cases also imply substantial changes in the future operations of U.S. petroleum refineries, as is particularly evident in the Low/No Net Imports case. Drastically reduced product consumption and increased nonpetroleum sources of transportation fuels, taken in isolation, would tend to reduce utilization of U.S. refineries. The combination of higher domestic crude supply and reduced crude runs in the refining sector would sharply reduce or eliminate crude oil imports and could potentially create market pressure for crude oil exports to balance crude supply with refinery runs. However, under current laws and regulations, crude exports require licenses that have not been issued except in circumstances involving exports to Canada or exports of limited quantities of specific crude streams, such as California heavy oil .
Rather than assuming a change in current policies toward crude oil exports, and recognizing the high efficiency and low operating costs of U.S. refineries relative to global competitors in the refining sector, exports of petroleum products, which are not subject to export licensing requirements, rise significantly to avoid the uneconomical unloading of efficient U.S. refinery capacity, continuing a trend that has already become evident over the past several years. Product exports rise until the incremental refining value of crude oil processed is equivalent to the cost of crude imports. To balance the rest of the world as a result of increased U.S. product exports, it is assumed that the increased volumes of U.S. liquid fuel product exports would result in a decrease in the volume of the rest of the world's crude runs, and that world consumption, net of U.S. exports, would also be reduced by an amount necessary to keep demand and supply volumes in balance.
Projected carbon dioxide emissions
Total U.S. CO2 emissions show the impacts of changing fuel prices through all the sectors of the economy. In the High Oil and Gas Resource case, the availability of more natural gas at lower prices encourages the electric power sector to increase its reliance on natural gas for electricity generation. Coal is the most affected, with coal displaced over the first part of the projection, and new renewable generation sources also affected after 2030 or so, resulting in projected CO2 emissions in the High Oil and Gas Resource case that exceed those in the Reference case after 2035 (Figure 25). With less-plentiful and more-expensive natural gas in the Low Oil and Gas Resource and High Net Imports cases, the reverse is true, with fewer coal retirements leading to higher CO2 emissions than in the Reference case early in the projection period. Later in the projection, however, the electric power sector turns first to renewable technologies earlier in the Low Oil and Gas Resource and High Net Imports cases, and after 2030 invests in more nuclear plants, reducing CO2 emissions from the levels projected in the Reference case. In the Low Oil and Gas Resource case, CO2 emissions are lower than in the Reference case starting in 2026. In the Low/No Net Imports case, annual CO2 emissions from the transportation sector continue to decline as a result of reduced travel demand; these emissions are conversely higher in the High Net Imports case. Figure 25 summarizes the CO2 emissions projections in the cases completed for this analysis.
NGL include a wide range of components produced during natural gas processing and petroleum refining. As natural gas production in recent years has grown dramatically, there has been a concurrent rapid increase in NGL production. NGL include ethane, propane, normal butane (n-butane), isobutane, and pentanes plus. The rising supply of some NGL components (particularly ethane and propane) has led to challenges, in finding markets and building the infrastructure necessary to move NGL to the new domestic demand and export markets. This discussion examines recent changes in U.S. NGL markets and how they might evolve under several scenarios. The future disposition of U.S. NGL supplies, particularly in international markets, is also discussed.
Recent growth in NGL production (Figure 39) has resulted largely from strong growth in shale gas production. The lightest NGL components, ethane and propane, account for most of the growth in NGL supply between 2008 and 2012. With the exception of propane, the main source of NGL is natural gas processing associated with growing natural gas production. That growth has led to logistical problems in some areas. For example, much of the increased ethane supply in the Marcellus region is stranded because of the distance from petrochemical markets in the Gulf Coast area.
The uses of NGL are diverse. The lightest NGL component, ethane, is used almost exclusively as a petrochemical feedstock to produce ethylene, which in turn is a basic building block for plastics, packaging materials, and other consumer products. A limited amount of ethane can be left in the natural gas stream (ethane rejection) if the value of ethane sinks too close to the value of dry natural gas, but the amount of ethane mixed in dry natural gas is small. Propane is the most versatile NGL component, with applications ranging from residential heating, to transportation fuel for forklifts, to petrochemical feedstock for propylene and ethylene production (nearly one-half of all propane use in the United States is as petrochemical feedstock). Butanes are produced in much smaller quantities and are used mostly in refining (for gasoline blending or alkylation) or as chemical feedstock. The heaviest liquids, known as pentanes plus, are used as ethanol denaturant, blendstock for gasoline, chemical feedstock, and, more recently, as diluent for the extraction and pipeline movement of heavy crude oils from Canada.
Unlike the other NGL components, a large proportion of propane is produced in refineries (which is mixed with refinery-marketed propylene). Given that refinery production of propane and propylene has been largely unchanged since 2005 at about 540 thousand barrels per day, the growth of propane/propylene supply shown in Figure 39 is solely a result of increased propane yields from natural gas processing plants.
International demand for NGL has provided an outlet for growing domestic production, and after years of being a net importer, the United States became a net exporter of propane in 2012 (Figure 40). Although the quantities shown in Figure 40, based on EIA data, represent an aggregated mixture of propane and propylene, other sources indicate that U.S. propylene exports have been on the decline since 2007 , implying that the recent change to net exporter status is the result of increased supplies of propane from natural gas processing plants.
Current developments in NGL markets
The market currently is reacting to the growing supply of ethane and propane by expanding both domestic use of NGL and export capacity. On the domestic side, much of the U.S. petrochemical industry can absorb ethane and propane by switching from heavier petroleum-based naphtha feedstock in ethylene crackers to lighter feedstock, and recent record low NGL prices have motivated petrochemical companies to maximize the amount of ethane and propane in their feedstock slate. To take advantage of the expected growth in supplies of light NGL components resulting from shale gas production, multiple projects and expansions of petrochemical crackers have been announced (Table 7).
Although the proposed projects shown in Table 7 will largely take advantage of the growing ethane supply, a few petrochemical projects that will use propane directly as a propylene feedstock through propane dehydrogenation also have been announced . Although expanded feedstock use is expected to be by far the largest source of expanded demand for NGL, increased use of NGL as a fuel, especially propane, also is expected—including the marketing of propane as an alternative vehicle fuel  and for agricultural use, with propane suppliers currently offering incentives for farmers to use propane as a fuel to power irrigation systems .
Notwithstanding the efforts to encourage the use of propane as a fuel in the United States, and despite current low prices, opportunities to expand the market for propane in uses other than as feedstock are limited. Therefore, producers, gas processors, and fractionators are looking for a growing export outlet for both ethane and liquefied petroleum gases (LPG—a mixture of propane and butane). Export capacity is being expanded, both on the U.S. Gulf Coast (Targa's expansion of both its gas processing and fractionation capability at Mont Belvieu and its export facility at Galena Park ) and on the U.S. East Coast (Sunoco Logistics' Mariner East project to supply propane and ethane to Philadelphia's Marcus Hook terminal [115, 116]). Exports of ethane from the Marcellus shale to chemical facilities in Sarnia, Ontario, via the Mariner West pipeline system, and from the Bakken formation to a NOVA Chemical plant near Joffre, Alberta, via the Vantage pipeline , are expected by the end of 2013. In addition to planned exports to Canada, a pipeline is being developed to transport ethane from the Marcellus to the Gulf Coast to relieve oversupply. The midstream sector's rapid buildup and expansion of natural gas processing, pipeline, and storage capacity have accommodated increasing volumes of NGL resulting from the sharp growth in shale gas production.
AEO2013 projects continued growth in both natural gas production and NGL supplies, with NGL prices determined in large part by Brent crude oil prices and Henry Hub spot prices for natural gas (Figure 41). In the AEO2013 Reference, Low Oil and Gas Resource, and High Oil and Gas Resource cases, industrial propane prices in 2040 range from $22.13 per million Btu (2011 dollars) in the High Oil and Gas Resource case to $27.48 per million Btu in the Low Oil and Gas Resource case, a difference of approximately 24 percent. The difference between the propane prices in the High and Low Oil and Gas Resource cases increases from $3.49 per million Btu in 2015 to $7.00 per million Btu in 2025 as natural gas prices and NGL production diverge in the two cases. Over time, however, as the divergence in NGL production narrows between the cases, the influence of oil prices on propane prices increases, and the difference in the propane prices narrows in the cases.
Production of NGPL, which are extracted from wet natural gas by gas processors, rises more steeply than natural gas production in the first half of the projection period as a result of increased natural gas and oil production from shale wells, which have relatively high liquids contents. As shale gas plays mature, NGPL production levels off or declines even as dry natural gas production increases (Figure 42).
Variations in NGL supplies and prices contribute to variations in demand for NGL. In the High Oil and Gas Resource case, propane demand in all sectors is higher than projected in the Reference case, and in the Low Oil and Gas Resource case propane demand is lower than in the Reference case. Some of the difference results from changes in the expected energy efficiency of space heating equipment in the residential sector, and possibly some fuel switching, in response to different price levels in the three cases. The remainder is attributed to variations in NGL feedstock consumption in the bulk chemicals sector, where the use of NGL as a fuel and feedstock varies with different price levels. In addition, because NGL feedstock competes with petroleum naphtha in the petrochemical industry, lower NGL prices relative to oil prices lead to more NGL consumption in the petrochemical industry.
The LPG import-export balance changes rapidly when domestic supply exceeds demand. This trend continues in the near term in all three cases. In the High Oil and Gas Resource case, however, with more LPG production, net exports continue to grow throughout the projection (Figure 43). Propane accounts for most of the higher export volumes, which also include smaller amounts of butane and ethane. Currently, most U.S. exports of LPG go to Latin America, where LPG is used for heating and cooking.
The projected growth in NGL demand both for U.S. domestic uses and for export depends heavily on international markets. Current plans for ethane exports are limited to pipelines to Canada, and to date ethane is not shipped by ocean-going vessels. There is room for growth in propane exports, however, because propane is a far more versatile fuel. Propane exports to Latin America are expected to continue, along with some expansion into European markets. In addition, growing markets in Africa  for propane used in heating and cooking, along with continued demand from Asia (for fuel and feedstock), are expected to support exports of propane from both the United States and the Middle East. It remains to be seen how the market for propane exports will develop in the long term, and how the United States will seek value for its propane—converting it into chemicals for domestic use or for export, or exporting raw propane.
International markets also play a role in increased domestic consumption, particularly for expanded petrochemical feedstock consumption. The declining price of ethane improves the economics of ethylene crackers, as indicated by the planned capacities shown in Table 7. The new capacity suggests that companies are planning to gain a greater market share of ethylene demand in Asia, especially in China, which continues to be a growing importer of ethylene . However, that economic advantage has to be weighed against the massive growth in chemical manufacturing complexes in the Middle East, as well as expansions in Asia. Feedstock availability will not be a concern in the Middle East, but most petrochemical plants in China and other Asian countries rely heavily on naphtha as a feedstock, and naphtha is produced from crude oil, which China imports. China is making efforts to diversify its feedstock slate and has announced plans to build coal-to-olefins plants . In addition, China may develop its own shale gas resources over the next 10 to 15 years, which could provide less expensive supplies of ethane and propane. The advantage in the Middle East is its long-term access to feedstocks. Whether the United States can further capitalize on growth in basic chemical production (ethylene, propylene) to build up its higher-value chemical base, and how the production cost of those higher value chemicals would compete with those from Asia and the Middle East, is an open question.
Future plans for U.S. propane disposition will be based on the balance between growth in domestic demand and exports. Rising exports of propane and butane raise issues as well. For example, both propane and butane can be used not only as feedstock in ethylene crackers, but also as feedstock for specific chemical product. For example, dehydrogenation processes can make propylene from propane  and butadiene from butane . The economic value of those chemicals (which would depend on both local and global markets), weighed against the export value of the NGL inputs (propane and butane), will need to be assessed. In addition, the value of derivatives (such as polyethylene and polypropylene) will be considered from the perspective of both their export value and their production costs, which will be tied directly to the price of their precursor inputs, ethylene and propylene. Finally, U.S. refineries produce a significant amount of propylene. There is some degree of flexibility within refineries' fluid catalytic cracker units to produce propylene , and future refinery production of propylene will depend on the value of propylene itself, the value of its co-products (mostly gasoline and propane), and refining costs.
Endnotes for Issues in Focus
70. Volatility is a measure of variability in a data series over time (more technically, the annualized standard deviation from the mean). This analysis was conducted using the GARCH estimation method for monthly average Brent crude oil prices.
71. Liquid fuels consists of crude oil and condensate to petroleum refineries, refinery gain, NGPL, biofuels, and other liquid fuels produced from non-crude oil feedstocks such as CTL and GTL.
72. Geologic characteristics relevant for hydrocarbon extraction include depth, thickness, porosity, carbon content, pore pressure, clay content, thermal maturity, and water content.
73. A production type curve represents the expected production each year from a well. A wellâ€™s EUR equals the cumulative production of that well over a 30-year productive life, using current technology without consideration of economic or operating conditions. A description of a production type curve is provided in the Annual Energy Outlook 2012 "Issues in focus" article, "U.S. crude oil and natural gas resource uncertainty," http://www.eia.gov/forecasts/archive/aeo12/IF_all.cfm#uscrude.
74. A more detailed analysis of the uncertainty in offshore resources is presented in the Annual Energy Outlook 2011 "Issues in focus" article, "Potential of offshore crude oil and natural gas resources," http://www.eia.gov/forecasts/archive/aeo11/IF_all.cfm#potentialoffshore.
75. U.S. Environmental Protection Agency and National Highway Transportation Safety Administration, "2017 and Later Model Year Light-Duty Vehicle Greenhouse Gas Emissions and Corporate Average Fuel Economy Standards: Final Rule," Federal Register, Vol. 77, No. 199 (Washington, DC: October 15, 2012), https://www.federalregister.gov/articles/2012/10/15/2012-21972/2017-and-later-model-year-light-duty-vehicle-greenhouse-gas-emissions-and-corporate-average-fuel.
76. K.A. Small and K.Van Dender, "Fuel Efficiency and Motor Vehicle Travel: The Declining Rebound Effect," University of California, Irvine, Department of Economics, Working Paper #05-06-03 (Irvine, CA: August 18, 2007), http://www.economics.uci.edu/files/economics/docs/workingpapers/2005-06/Small-03.pdf.
77. National Petroleum Council, "Advancing Technology for Americaâ€™s Transportation Future" (Washington, DC: August 1, 2012), http://www.npc.org/FTF-report-080112/NPC-Fuels_Summary_Report.pdf.
78. International Maritime Organization, Information Resources on Air Pollution and Greenhouse Gas (GHG) Emissions from International Shipping (Marpol Annex VI (SOX, NOX, ODS, VOC) / Greenhouse Gas (CO2) and Climate Change) (London, United Kingdom: December 23, 2011), http://www.imo.org/KnowledgeCentre/InformationResourcesOnCurrentTopics/AirPollutionand
79. U.S. Energy Information Administration, Could the United States become the leading global producer of liquid fuels, and how much does it matter to U.S. and world energy markets?," This Week in Petroleum (Washington, DC: December 19, 2012), http://www.eia.gov/oog/info/twip/twiparch/2012/121219/twipprint.html.
80. U.S. Energy Information Administration, "Could the United States become the leading global producer of liquid fuels, and how much does it matter to U.S. and world energy markets?," This Week in Petroleum (Washington, DC: December 19, 2012), http://www.eia.gov/oog/info/twip/twiparch/2012/121219/twipprint.html.
81. The circumstances under which the United States can and cannot export crude oil under current law are more fully described in U.S. Energy Information Administration, "Market implications of increased domestic production of light sweet crude oil?," This Week in Petroleum (Washington, DC: November 28, 2012), http://www.eia.gov/oog/info/twip/twiparch/2012/121128/twipprint.html.
110. Global Data, "Propylene Exports," Petrochemicals eTrack (March 2013), http://petrochemicalsetrack.com (subscription site).
111. A. Greenwood, "Tight US propylene may lead to two more PDH plants," ICIS News (June 25, 2012), http://www.icis.com/Articles/2012/06/25/9572490/tight-us-propylene-may-lead-to-two-more-pdh-plants.html.
112. J. Schroeder, "ProCOT Launches Propane Education Campaign," Ethanol Report (March 11, 2013), http://domesticfuel.com/2013/03/11/procot-launches-propane-education-campaign.
113. Propane Education and Research Council, "Programs and Incentives," http://www.agpropane.com/programs-and-incentives.
114. Targa Resources, "Targa Resources Partners LP Announces Project to Export International Grade Propane," Globe Newswire (Houston, TX: September 19, 2011), http://ir.targaresources.com/releasedetail.cfm?releaseid=606206.
115. A. Maykuth, "Former Sunoco Refinery in Marcus Hook Will Process Marcellus Shale Products," The Philadelphia Inquirer (Philadelphia, PA: September 28, 2012), http://articles.philly.com/2012-09-28/business/34128480_1_ethane-sunoco-logistics-sunoco-s-marcus-hook.
116. "More US Marcellus projects needed to absorb ethane production," ICIS News (Houston, TX: October 10, 2011), http://www.icis.com/Articles/2011/10/10/9498866/more-us-marcellus-projects-needed-to-absorb-ethane-production.html.
117. C.E. Smith, "US NGL Pipelines Expand to Match Liquids Growth," Oil and Gas Journal (May 7, 2012), http://www.ogj.com/articles/print/vol-110/issue-5/special-report-worldwide-gas/us-ngl-pipelines-expand.html (subscription site).
118. S. Williams, "LPG - the fuel for Africa," New African (December 2007), http://www.africasia.com/uploads/na_oilgas_1207.pdf.
119. "China's ethylene import dependency rises in January-May," ICIS C1Energy (July 17, 2012), http://www.c1energy.com/common/4162279,0,0,0,2.htm.
120. B. Thiennes, "Increased Coal-to-Olefins Processes in China," Hydrocarbon Processing (Houston, TX: October 1, 2012), http://www.hydrocarbonprocessing.com/Article/3096211/Increased-coal-to-olefins-processes-in-China.html.
121. A. Greenwood, "Tight US propylene may lead to two more PDH plants," ICIS News (June 25, 2012), http://www.icis.com/Articles/2012/06/25/9572490/tight-us-propylene-may-lead-to-two-more-pdh-plants.html.
122. J. Richardson, "Butadiene Oversupply Threat," ICIS Asian Chemical Connections (May 10, 2012), http://www.icis.com/blogs/asian-chemical-connections/2012/05/butadiene-oversupply-threat.html.
123. K.A. Couch et. al., "FCC propylene production," Petroleum Technology Quarterly (2007), http://www.uop.com/wp-content/uploads/2011/02/UOP-FCC-Propylene-Production-Tech-Paper.pdf.
139. One option for balancing the mix of crudes might be to allow the export of domestically produced light crude in exchange for heavier crudes. Crude exports and swaps, however, are currently permitted only in limited cases and require a license from the Department of Commerce.
- Oil production, particularly from tight oil plays, rises over the next decade, leading to a reduction in net import dependence
- Natural gas consumption grows in industrial and electric power sectors as domestic production also serves an expanding export market
- Range of oil price cases represents uncertainty in world oil markets
- Trends in petroleum and other liquids markets are defined largely by the developing nations
- Production of liquid fuels from biomass, coal, and nagural gas increases
- With production outpacing consumption, U.S. exports of natural gas exceed imports
- Pipeline exports increases ad Canadian imports fall and exports to Mexico rise
- Domestic production of tight oil leads to lower imports of light sweet crude oil
- Increasing U.S. supply results in decreasing net imports of petroleum and other liquids
Issues in Focus